Amazon Web Services has rolled out Amazon Bio Discovery, a no-code AI tool that speeds up early-stage drug research. Scientists pick from a library of biological foundation models to generate and score new molecules. An onboard AI agent guides model selection, parameter settings and result interpretation. Once candidates clear computational screening, they’re sent to lab partners—like Twist Bioscience in a Memorial Sloan Kettering collaboration—where nearly 300,000 AI-designed antibodies were winnowed down to 100,000 for physical testing, cutting months of work into weeks. Big names have already signed on. Bayer, the Broad Institute and Voyager Therapeutics are among early adopters. In fact, 19 of the top 20 global pharma firms already use AWS cloud services. Rajiv Chopra, AWS’s vice president for healthcare AI, says the main bottleneck today is translating lab questions into machine-learning pipelines. The service is meant to assist scientists and contract research organizations, not replace them. Outside drug design, AWS, Boston Consulting Group and Merck are teaming up on another AI platform to improve clinical trial site selection, a frequent hold-up in taking new drugs from lab to market. Jefferies analyst Tycho Peterson notes that fears of AI cutting demand for research instruments are overblown; higher R&D returns could spur more spending on these tools.